基于局部和全局知识蒸馏的危险驾驶行为检测

doi:10.6040/j.issn.1672-3961.0.2024.172

摘要/Abstract

摘要： 为提高道路安全,预防交通事故,提出一种基于局部与全局知识蒸馏(local and global knowledge distillation, LGD)的危险驾驶行为检测算法。在知识蒸馏框架基础上,引入融合局部特征与全局特征的蒸馏损失函数,引导学生网络高效学习教师模型判别能力。这一方法有效促进学生网络的学习,使其能够在保持参数减少时,达到与教师网络相当的检测准确率。试验结果显示,本方法在仅包含31.85 M参数大小的模型下,实现91.79%准确率,表明该方法在处理分心驾驶检测问题上的有效性。

关键词: 知识蒸馏, 学生网络, 分心驾驶检测

Abstract: To enhance road safety and prevent traffic accidents, a distracted driving behavior detection algorithm based on local and global knowledge distillation(LGD)was proposed. Built upon a knowledge distillation framework, the method introduced a distillation loss function that integrated both local and global features, guiding the student network to effectively learn the discriminative capabilities of the teacher model. While maintaining a lightweight network structure, the approach significantly improved the recognition accuracy of distracted driving behaviors. By effectively facilitating the learning process of the student network, the method enabled it to achieve detection accuracy comparable to that of the teacher model, despite having fewer parameters. Experimental results demonstrated that the proposed method achieved an accuracy of 91.79% with only 31.85 M parameters, highlighting its effectiveness in addressing the problem of distracted driving detection.

Key words: knowledge distillation, student network, distracted driving detection

中图分类号:

TP391

李坤彪,杨晓晖,张风,徐涛,郭庆北. 基于局部和全局知识蒸馏的危险驾驶行为检测[J]. 山东大学学报 (工学版), 2025, 55(6): 13-20.

LI Kunbiao, YANG Xiaohui, ZHANG Feng, XU Tao, GUO Qingbei. Risky driving behavior detection based on local and global knowledge distillation[J]. Journal of Shandong University(Engineering Science), 2025, 55(6): 13-20.

参考文献

[1] World Health Organization. Road traffic injuries[EB/OL].(2023-12-13)[2024-07-17]. https://www.who.int/newsroom/factsheets/detail/road-traffic-injuries
[2] NASR AZADANI M, BOUKERCHE A. Driving behavior analysis guidelines for intelligent transportation systems[J]. IEEE Transactions on Intelligent Transportation Systems, 2021, 23(7): 6027-6045.
[3] WANG J Y, CHAI W H, VENKATACHALAPATHY A, et al. A survey on driver behavior analysis from in-vehicle cameras[J]. IEEE Transactions on Intelligent Transportation Systems, 2021, 23(8): 10186-10209.
[4] NHTSA. Put the phone away or pay[EB/OL].(2024-04-01)[2024-07-17]. https://www.nhtsa.gov/press-releases/2022-traffic-deaths-2023-early-estimates
[5] U. S. Center for Disease Control and Prevention. Distracted driving[EB/OL].(2024-05-16)[2024-07-17]. https://www.cdc.gov/distracted_driving/about/
[6] ZANGI N, SROUR-ZREIK R, RIDEL D, et al. Driver distraction and its effects on partially automated driving performance: a driving simulator study among young-experienced drivers[J]. Accident Analysis & Prevention, 2022, 166: 106565.
[7] HANG P, LV C, XING Y, et al. Human-like decision making for autonomous driving: a noncooperative game theoretic approach[J]. IEEE Transactions on Intelligent Transportation Systems, 2020, 22(4): 2076-2087.
[8] JHA S, BUSSO C. Estimation of driver's gaze region from head position and orientation using probabilistic confidence regions[J]. IEEE Transactions on Intelligent Vehicles, 2022, 8(1): 59-72.
[9] ROY K. Unsupervised sparse, nonnegative, low rank dictionary learning for detection of driver cell phone usage[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(10): 18200-18209.
[10] ZUO X, ZHANG C, CONG F Y, et al. Mobile phone use driver distraction detection based on MSaE of multi-modality physiological signals[J]. IEEE Transactions on Intelligent Transportation Systems, 2024, 25(11): 17650-17665.
[11] GIRSHICK R, DONAHUE J, DARRELL T, et al. Region-based convolutional networks for accurate object detection and segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 38(1): 142-158.
[12] REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2016, 39(6): 1137-1149.
[13] CAI Z W, VASCONCELOS N. Cascade R-CNN: delving into high quality object detection[C] // Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, USA: IEEE, 2018: 6154-6162.
[14] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C] //Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, USA: IEEE, 2016: 779-788.
[15] REDMON J, FARHADI A. YOLOv3: an incremental improvement[EB/OL].(2018-04-08)[2024-07-17]. https://arxiv.org/abs/1804.02767
[16] TIAN Z, SHEN C H, CHEN H, et al. FCOS: fully convolutional one-stage object detection[C] //Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019:9626-9635.
[17] HINTON G, VINYALS O, DEAN J. Distilling the knowledge in a neural network[J]. Computer Science, 2015, 14(7):38-39.
[18] ROMERO A, BALLAS N, KAHOU S E, et al. Fitnets: hints for thin deep nets[EB/OL].(2014-12-19)[2024-07-17]. https://arxiv.org/abs/1412.6550
[19] ZAGORUYKO S, KOMODAKIS N. Paying more attention to attention: improving the performance of convolutional neural networks via attention transfer[EB/OL].(2016-12-12)[2024-07-17]. https://arxiv.org/abs/1612.03928
[20] CHEN G B, CHOI W G, YU X, et al. Learning efficient object detection models with knowledge distillation[C] //Proceedings of the Advances in neural information processing systems. Long Beach, USA: NIPS, 2017:742-751.
[21] LI Q Q, JIN S Y, YAN J J. Mimicking very efficient network for object detection[C] //Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, USA: IEEE, 2017: 7341-7349.
[22] ZHENG Z H, YE R G, HOU Q B, et al. Localization distillation for object detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45(8): 10070-10083.
[23] YANG Z D, LI Z, JIANG X H, et al. Focal and global knowledge distillation for detectors[C] //Proceedings of 2022 the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans, USA: IEEE, 2022: 4633-4642.
[24] STREIFFER C, RAGHAVENDRA R, BENSON T, et al. Darnet: a deep learning solution for distracted driving detection[C] //Proceedings of the 18th ACM/IFIP/USENIX Middleware Conference Industrial Track. Las Vegas, USA: ACM, 2017: 22-28.
[25] LE T H N,ZHENG Y T, ZHU C C, et al. Multiple scale faster-RCNN approach to driver's cell-phone usage and hands on steering wheel detection[C] //Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition Workshops. Las Vegas, USA: IEEE, 2016: 46-53.
[26] JAIN H, ZOPE H, JAIN H, et al. Mobile phone detection and alcohol detection with engine locking for the prevention of car accidents[C] //Proceedings of the 2023 International Conference on Digital Applications, Transformation & Economy. Miri, Malaysia: IEEE, 2023: 1-6.
[27] ABOUELNAGA Y, ERAQI H M, MOUSTAFA M N. Real-time distracted driver posture classification[EB/OL].(2017-06-28)[2024-07-17]. https://arxiv.org/abs/1706.0948
[28] WAGNER B, TAFFNER F, KARACA S, et al. Vision based detection of driver cell phone usage and food consumption[J]. IEEE Transactions on Intelligent Transportation Systems, 2021, 23(5): 4257-4266.
[29] JIANG L D, XIE W, ZHANG D, et al. Smart diagnosis: deep learning boosted driver inattention detection and abnormal driving prediction[J]. IEEE Internet of Things Journal, 2021, 9(6): 4076-4089.
[30] CAI J F, BAI J J, ZHOU T Q, et al. Dangerous driving behavior detection based on multi-source information fusion[C] //Proceedings of the 2022 8th International Conference on Big Data and Information Analytics(BigDIA), Guiyang, China: IEEE, 2022: 366-370.
[31] NARAYANA P, ATTAR N. Analyzing the impact of distractions on driver attention: insights from eye movement behaviors in a driving simulator[C] //Proceedings of the 2023 Seventh IEEE International Conference on Robotic Computing(IRC). Laguna Hills, USA: IEEE, 2023: 356-359.
[32] LIU R, QI G Q, GUAN W, et al. Driver distraction state recognition based on minimum spanning tree features using EEG data[C] //Proceedings of the 2023 3rd International Conference on Digital Society and Intelligent Systems(DSInS). Chengdu, China: IEEE, 2023: 437-440.
[33] BACHTIAR F A, ARWANI I, PRASETYO R A B, et al. Distraction detection in driving using pose extraction and machine learning techniques[C] //Processdings of the 2023 IEEE 7th International Conference on Information Technology, Information Systems and Electrical Engineering(ICITISEE). Purwokerto, Indonesia: IEEE, 2023: 279-284.
[34] PUTRA N C B, YUNIARNO E M, RACHMADI R F. Driver visual distraction detection based on face mesh feature using deep learning[C] // 2023 International Seminar on Intelligent Technology and Its Applications(ISITIA). Surabaya, Indonesia: IEEE, 2023: 6-11.
[35] CAO Y, XU J R, LIN S, et al. GCNet: non-local networks meet squeeze-excitation networks and beyond[C] //Processdings of the 2019 IEEE/CVF International Conference on Computer Vision Workshops. Seoul: IEEE, 2019: 1971-1980.
[36] LI X, WANG W H, WU L J, et al. Generalized focal loss: Learning qualified and distributed bounding boxes for dense object detection[C] //Proceedings of the 34th International Conference on Neural Information Processing Systems. Vancouver, Canada: ACM, 2020: 21002-21012.
[37] YANG Z, LIU S H, HU H, et al. RepPoints: point set representation for object detection[C] //Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019: 9656-9665.
[38] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C] //Proceedings of the 2017 IEEE International Conference on Computer Vision. Venice, Italy: IEEE, 2017: 2999-3007.

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed